Did Intel only Release the 11th generation to Interfere with AMD’s growing Marketshare

[the wopster]

TO BE CLEAR THIS THREAD IS NOT A THREAD TO ARGUE WHO IS THE BEST AND IS FOR INTELLIGENT DISCUSSION ONLY!!!!

Considering Intel have been market leaders for well over a decade and have almost become unaccustomed to almost any form of challenge/competition, was the release of the 11th Gen CPU family more of a ploy to try and disrupt AMDs growing marketshare in the mid range, high end and server areas, than to release something that’s an upgrade to 10th gen or a direct challenge Ryzen/Threadripper/Epic?


For clarification I have no affiliation to Intel or AMD I have build and owned numerous systems powered by both Processors.

 

[Journey]

The decision to backport it obviously happened a good while ago now, fundamentally the reason to release it is continued progression, but I guess they couldn't see how little progress would be made until a great deal of the work was done and the R&D money already spent. Why not try and recoup some of that money, using it as a marketing exercise, and get PCI-E 4.0 out of the door to the masses before it is redundant. Obviously they - want - to compete with AMD, but at the same time they need to ensure that they appease their shareholders. I think we can see with the change of CEO, that the decisions made that allowed this to happen have now been dealt with, and moving forward I expect a refreshed and renewed Intel, bring on 2023!

 

[the wopster]

The decision to backport it obviously happened a good while ago now, fundamentally the reason to release it is continued progression, but I guess they couldn't see how little progress would be made until a great deal of the work was done and the R&D money already spent. Why not try and recoup some of that money, using it as a marketing exercise, and get PCI-E 4.0 out of the door to the masses before it is redundant. Obviously they - want - to compete with AMD, but at the same time they need to ensure that they appease their shareholders. I think we can see with the change of CEO, that the decisions made that allowed this to happen have now been dealt with, and moving forward I expect a refreshed and renewed Intel, bring on 2023!

All of that is totally understandable if they had nothing until mid/late 2022, but isn’t Alder Lake coming towards the end of this year, there is no doubt they didn’t expect AMD to be a threat and AMD did catch them with their pants down
That’s why I asked if it was more of a disruptive tactic to stop an AMD foothold into a higher marketshare until they got something out that is truly competitive on all levels, they still have no answer for the 5950x and have taken away the counter to the 39/5900x by taking away the 10 core chip(yes I know they die would have been to big).


Alder Lake looks very interesting, 8 big and 8 little cores is a very interesting concept

 

[Perfect_Chaos]

Alder Lake looks very interesting, 8 big and 8 little cores is a very interesting concept

Depends how powerful the big cores will be. I was watching the Tech Deals guy yesterday and he was saying that LGA1700 might support a few CPU generations.
But i would not trust that to mean buying a board will mean you can upgrade to a Meteor Lake chip a couple of years later.

 

[the wopster]

Depends how powerful the big cores will be. I was watching the Tech Deals guy yesterday and he was saying that LGA1700 might support a few CPU generations.
But i would not trust that to mean buying a board will mean you can upgrade to a Meteor Lake chip a couple of years later.

Normally 2 generations per platform, I wouldn’t expect any change on that front TBH

 

[KompuKare]

Alder Lake looks very interesting, 8 big and 8 little cores is a very interesting concept

"May you live in interesting times".
Which may not be a Chinese proverb / curse, but maybe Intel are going to say something like this all kernel developers.
It certainly is a bold move in the x86 even if they are just copying ARM's idea.
Isn't the rumour that Intel are going to add their own hardware scheduler to it, so that would imply lots of microcode updates to get it all working well.
As for Rocket Lake being a Zen 3 spoiler?
While that may be what has happened to some extent (but being behind Zen3 at double the power draw isn't much of a spoiler), those decisions would all have had to be made years ago. Certainly the back port go-ahead was probably 2 years ago. Market position and final clocks would have been much more recent. The i9 is a complete mystery as I can't see objective person buying them.

 

[Destination]

They need to release something, and talk works, they need to release as the shareholders wouldn't stomach a missed launch.
Alderlake we keep being told is 6 months away, let us see if that occurs, it might, be until it is here, we can't judge it.
If their stock forecasts say they will release a major line on x, then it needs tog et out there, else it scuppers their stock price, and class action lawsuits happen.

They should have just rebadged the 10 core, then released the 8 and 6 cores to be competitive, thye didn't.
Made the 11600 the 11400, and at that price.
made the 11700 the 11600, and at that price.
Had the 11900 as a refresh on the 10900, and slashed prices dramatically.

Then they'd have had an eyebrow raising competitive launch.
They don't care, they ae too big, they just needed to push any old ***** out the door, and it'll do.
It'll take several years of 'it'll do' before it become an issue.

I hope alderlake burns and does tremendously badly, so the progression of ARM and AMD actually affords us a positive and completely different way forward.
It'll force intel to actually bother, and that will lead to 5 and 10 year progression.

I moved from a i7 920 to the 6 core xeon of the same ilk (£15) before eventually jumping to a 3600, which I will likely make into a 5900x eventually.
The jump to 3600 doubled my perf. It'll do the same with a 5900 upgrade.
This I found a very acceptable upgrade after many many years, but the 4-6% of intel stage progression each year over a massive number of years is awful.
At least AMD seem to have got their act together, hopefully this continues.

 

[d_brennen]

Products in the channel = sales

Sales = profits

Profits = happy shareholders

 

[Jeps]

Good video from Dr Ian Cuttress about intels reasoning. Basically the design engineering experience is invaluable to intel's engineering teams, even if the product isn't necessarily a massive success vs the competition.

 

[Donnie Fisher]

I dont think it was just to disrupt market share, I think it was more along the lines of just having to release something during this cycle because that it what was expected by shareholders who care more about bottom line than the actual product. I think the pricing of it is poor for the 11900K ... it does seem to be somewhat of a cobbled together product, which doesn't really seem to offer another level of improvement over previous, so to ask more from it seems naive.

To me, the 11900K as a product should have released under cutting the 10 series chip price, highlighting the newer process being cheaper but about the same performance as before ( in certain tasks). that way it could have been marketed as an improvement.

Plus it tests a new design on an existing manufacuring process ... means you only concentrate on one aspect of it rather than both.

 

[KompuKare]

I dont think it was just to disrupt market share, I think it was more along the lines of just having to release something during this cycle because that it what was expected by shareholders who care more about bottom line than the actual product. I think the pricing of it is poor for the 11900K ... it does seem to be somewhat of a cobbled together product, which doesn't really seem to offer another level of improvement over previous, so to ask more from it seems naive.

To me, the 11900K as a product should have released under cutting the 10 series chip price, highlighting the newer process being cheaper but about the same performance as before ( in certain tasks). that way it could have been marketed as an improvement.

Plus it tests a new design on an existing manufacuring process ... means you only concentrate on one aspect of it rather than both.

Err?
Rocket Lake is the same 14nm as Comet Lake so no new process just (finally) their first new (desktop) architecture since Skylake still using 14nm first introduced with Broadlake. Although Intel have spent the last 6+ years adding +'s to their 14nm.

 

[Joxeon]

Intel have certainly disrupted AMD at the lower end where anyone who just games would have to be brain dead now to buy a 5600X when you can get an 11400F MB and ram for the same price as just the AMD CPU with virtually the same gaming performance outside of 1080p with a 3090.

 

[Grim5]

Intel have certainly disrupted AMD at the lower end where anyone who just games would have to be brain dead now to buy a 5600X when you can get an 11400F MB and ram for the same price as just the AMD CPU with virtually the same gaming performance outside of 1080p with a 3090.

No one buys the 3090 with an 11400f you'd be brain dead as you put it

 

[SKILL]

Wasn't the original plan, way back when, to release 7nm products this year? And presumably 10nm products (on the desktop) before that?

I don't think Rocket Lake was planned until relatively recently (not that recently, still takes time, obviously), and it is what it is... Better to get something than have nothing to 'compete' with Zen3, they've clearly sold, and will sell, chips and motherboards/chipsets off the back of it so it's not a complete loss. They've even managed to hit some market segments where they are competitve, notably the 11400/11600 chips, that alone likely makes it worth it.

I wouldn't be surprised if Alder Lake is one if not both of:

• Late (mid-2022 maybe?)
• Mediocre on desktop

There must be a reason why they've had '10nm' for years no without releasing a desktop chip, and I don't think it's capacity. With 7nm seemingly having issues/completely broken they need to get new chips out, 14nm is very clearly EOL at this point, but I don't think 10nm is in any way their ideal option.

 

[Grim5]

Don't expect meteor lake to be cheap according to Ian. Foveros EMIB has higher upside potential than AMD's infinity fabric but it's also more expensive

 

[Michael_T]

Rocket lake was meant to be on their new 10nm node, but they backported it because they couldn't get the 10nm process to work reliably. The decision to backport wouldn't have happened recently, as backporting it is a lot of work. Rocket lake, architecturally, is an update over Comet Lake and introduces PCIE 4.0 support. But since it still uses 14nm, it can't get the power consumption down, so performance is more or less the same as Comet Lake.

I suspect Intel released it because they wanted to get something out with the new architecture to compete feature wise with AMD. They will have known the performance wasn't there, but once they'd already done all that work, it obviously made more sense to release it than to scrap it - which is the only other option.

Rocket Lake will still sell, we're far more informed about this stuff than your average consumer and business customers will still be buying 11th Gen processors. Partners like HP, Dell etc, will still be updating their new product lines to use 11th Gen so business customers will still buy them. It may not be good value and might not be a sensible purchase if you pay attention to all the reviews, but we're not the target market for most of Intel's market share.

 

[humbug]

From another thread....

Margins, Intel know most people who buy thier CPU's will go for the 11600K and 11700K, at $270 and $400 they are not making as much money, or (Margins) as Intel would like, so the 11900K at $540 off sets that to keep the margins up, some people will buy them over the 11700K just because its the most expensive CPU they make, its E-Peen.

Its about making as much money as possible and keeping the margins on paper up, Intel's margins are going down because they have been forced to lower prices, Intel's investors don't like this, if they see Intel's margins go from 65% to 62% they have a panic attack.

Investors losing patience with Intel is probably why they put these things on the market in the first place, its to say "we are combating AMD" and then put out a bunch of marketing slides about how they are better than AMD at everything (which they did) and hope these investors buy it hook line and sinker, which often they do.

This decision was probably made about a year ago, if not 2 years.

I don't think Intel expected AMD to gain 20 to 45% in gaming over Zen 2 with Zen 3, Intel expected to retain the gaming crown and expected to be able to make a huge marketing woof about it, much to their frustration now i'm sure...

 

[g67575]

No one buys the 3090 with an 11400f you'd be brain dead as you put it

Braindead, but still relatively rich.

Tbh, might be ok for a year, just buy new mobo + CPU for alder lake (or, maybe zen 4 if it also supports ddr4 RAM).

 

[chrcoluk]

I think it was released because tech companies have got used to releasing on a schedule vs releasing when a product is ready, I dont like this new model but it is dominant across tech companies now. I am assuming its proven to be more profitable.

It is an interesting one as this gen of intel cpu has a decent IPC boost vs its previous gen, AMD seem to have pressured them into doing that, but the power usage is just out of this world, if it was same as say 9900k which is same core count I felt it would have been an ok cpu, not great but ok.

 

[g67575]

The difference is, AMDs claimed IPC gains vs previous generations actually count for something in games + software.

Intel plans to invest heavily in their chip foundries, I think they may catch up with AMD in 2023 (7nm EUV CPUs), and do reasonably well in 2022.

The fact that Intel are pushing towards 7nm EUV CPUs suggests to me that the transistor density of their 10nm desktop chips won't be enough to compete with 5nm EUV based Zen 4 CPUs, in 2022/2023.

I wonder if they will try to improve their 10nm process further for Alder Lake's successor, Raptor Lake?

Intel has publicly expressed concern regarding the improvements offered by 10nm, and their desire to release 7nm products as soon as possible.